Ethereum wants to cut staking rewards, traditional automated trading strategies face challenges

On August 4, six Ethereum researchers, including Justin Drake, submitted EIP-8363.

This proposal is named Tapered Issuance Burn, which literally translates to progressive issuance destruction. The name is technical, but the concept is easy to understand: the more ETH that enters staking, the more validator rewards Ethereum destroys.

Once about half of the ETH enters staking, the destroyed portion will completely offset the consensus layer issuance rewards received by validators.

Currently, about one-third of the ETH has entered staking. According to the long-term formula in the proposal, the current net consensus layer yield of approximately 2.6% may drop to around 1.2%, a decrease of more than half. Therefore, the proposal authors have set an 18-month transition period. (EIP-8363)

The proposal is still in draft stage, and there is a long process before it enters the mainnet. The market needs to start discussing it, as EIP-8363 touches on an important underlying parameter: Ethereum staking yield.

Many people may first react to this by thinking about how much less validators will earn in the future.

The impact goes far beyond that.

Staking yield has already been embedded in liquid staking, lending, leveraged strategies, and institutional ETH products. Once this number changes, a large number of strategies will need to be recalculated.

What’s more troublesome is that many automated trading systems may still be operating under the old formula.

Staking yield has become the base interest rate of Ethereum

Ethereum relies on validators to confirm transactions and maintain network operation.

Validators first deposit ETH and then run software to check new blocks. This ETH is similar to a performance bond. Completing the work can earn rewards, while serious violations will incur penalties.

Ethereum can be understood as a clearing company without a headquarters.

Validators are accountants distributed around the globe. The staked ETH serves as collateral. The consensus layer issuance rewards are equivalent to a base salary, while transaction fees and MEV are additional income.

Currently, about 41.5 million ETH are in a staked state, accounting for about one-third of the total supply. The annualized staking yield displayed on the page is approximately 2.6%. Running an independent validator requires 32 ETH, but users with less capital can also participate through staking pools.

Platforms like Lido and Rocket Pool also issue liquid staking tokens such as stETH and rETH to users. Users can earn staking rewards while continuing to use these tokens for trading and lending. (Ethereum Staking)

EIP-8363 is preparing to adjust the base income of validators.

Under the current rules, the more ETH that enters staking, the less reward each unit of capital receives. New staking can still earn positive returns. The proposal document believes that even if the staking rate rises to a very high level, the current curve will still retain a yield floor of about 1.5%.

The proposal authors are concerned that this will continue to drive ETH concentration towards large staking institutions.

In the past, running a validator required equipment, technology, and maintenance capabilities. Now, staking tools have matured, and custodial services are lowering the barriers to entry. Funds and exchange-traded products may also bring in more institutional capital.

As long as new staking can always earn issuance rewards, more ETH may continue to flow into exchanges, funds, and large staking platforms.

The solution provided by EIP-8363 is a gradually tightening yield curve.

When the staking rate is low, validators can retain most of the issuance rewards. As the staking rate rises, the proportion of rewards destroyed by the protocol increases. When the staking balance reaches 60.25 million ETH, the deduction ratio will reach 100%. This scale is close to half of the current total supply of ETH.

Transaction fees and MEV income are still retained.

This can be understood as a company hiring people based on a fixed workload.

When the team is small, hiring one more employee can significantly improve efficiency. Once the team size is sufficient, the help brought by new employees becomes smaller, while wage expenses continue to grow.

EIP-8363 attempts to weaken the economic incentive for continued expansion.

The problem is that staking yield has long since left the small circle of validators.

Liquid staking tokens are widely used for DeFi collateral. Some users will borrow stablecoins using stETH. Markets like Pendle will separate principal and future earnings, allowing investors to trade them separately. Funds holding staked ETH often do so to increase yield on their positions. (CoinDesk)

After the base yield decreases, the prices across the entire chain will need to be recalculated.

How much income stETH can generate in the future needs to be reassessed. Reasonable borrowing costs will change. The profit margin for leveraged staking may disappear. Institutional products will also need to adjust expected returns.

An ETH holder who has never run a validator may also feel the impact through lending rates and fund returns.

Staking yield is approaching the base interest rate in the Ethereum economy.

It cannot determine all prices, but it influences the calculation methods of many products.

The protocol begins to change the rules, while the bots may still be reading the old map

Traditional markets have long been accustomed to pricing rule changes.

Bond traders watch central banks, stock investors focus on regulations and taxes. The crypto market has an additional policymaker, which is the protocol itself.

Ethereum Improvement Proposal, abbreviated as EIP, is a public proposal for modifying Ethereum rules.

A draft-stage EIP is very similar to a newly submitted bill. Developers will discuss the design, and client teams will assess the technical feasibility. Core developers may then incorporate it into a network upgrade.

The rules only take effect once the software has completed development, testing, and activation on the mainnet. (Ethereum EIPs)

The market usually acts in advance.

Once a proposal begins to attract attention, traders will assess three questions: how likely it is to pass, when it might take effect, and which parameters will be retained.

These judgments may influence positions even before the code enters the mainnet.

Taking a common leveraged staking strategy as an example.

Users first deposit ETH into a liquid staking protocol and receive a token with yield. Then, they use this token as collateral to borrow stablecoins, and use the borrowed funds to purchase ETH and continue staking.

Assuming the staking yield is 2.6% and the borrowing cost is 2%. The profit margin is thin but still positive.

Once the staking yield drops to 1.2%, even if the ETH price remains unchanged, this strategy may begin to incur losses.

Bots that only look at prices may perceive the market as calm. The profit basis of the strategy has disappeared.

This is precisely where existing automated trading systems are prone to overlook.

Most trading bots operate around price, volume, and volatility. Developers write indicators and trigger conditions in advance, and the bots execute according to the rules.

Protocol parameters are often treated as fixed backgrounds.

Once the background conditions change, the models continue to run. The assumptions they use have expired.

EIP-8363 is also difficult to label as purely bullish or bearish.

A decrease in issuance can reduce the dilution faced by ETH holders, but it will also reduce the new ETH entering the market. A decline in validator earnings may drive some funds out of staking.

Independent validators may face greater pressure. Large institutions have stronger MEV acquisition capabilities and can more easily dilute equipment and operational costs.

A 2025 study on Ethereum staking found that independent stakers are more sensitive to changes in rewards. Its model suggests that a decrease in issuance rewards may crowd out some independent operators, increasing the market share of large institutions. (arXiv)

This means that the judgment that a decrease in issuance is bullish for ETH prices is far from sufficient.

Trading models need to continue to push forward.

How will validators act? Will staking become more concentrated? Will the demand for liquid staking decrease? How will borrowing rates change? How will the future yield market be repriced?

These changes rarely happen simultaneously.

Some impacts will be quickly priced in by the market, while others may take months to manifest.

If trading systems cannot update their assumptions, they may continue to execute after the market logic has already changed.

Next-generation trading systems need to update models before deciding whether to place orders

Most trading bots resemble a pre-printed map.

Developers select data, write opening rules, position limits, and exit conditions. The system can execute quickly, and then continue to use the same map until someone rewrites the program.

This method is effective as long as the roads remain unchanged.

Once the market begins to modify its own rules, the old map will quickly become obsolete.

A stronger agentic trading system should resemble real-time navigation.

When it discovers that a road may be closed, it will check the source of information, recalculate the route, and then decide whether to act based on user constraints.

In trading, this means updating the model first and then adjusting positions.

In the face of EIP-8363, such systems need to continuously monitor proposal documents, developer discussions, upgrade plans, and on-chain staking data, and then translate the information into different scenarios.

One scenario assumes the proposal ultimately fails.

Another scenario adopts an 18-month transition period.

The system can also test lower destruction ratios or later activation times.

Subsequently, it needs to recalculate expected yields, financing costs, and liquidation risks, and run historical tests using new assumptions.

If the results still meet the user's return targets and loss limits, the system will then decide whether to reduce leverage, establish hedges, or maintain positions.

Similar methods already exist.

After changes in central bank interest rate expectations, bond funds will rebuild interest rate models. On-chain yield markets will price future income separately. Lending protocols will also adjust collateral rules based on market changes.

Agentic trading can integrate these processes into a continuous workflow.

One user can allow the agent to automatically complete minor rebalancing while requiring all leverage adjustments to be approved first.

Another user can open analysis and simulation while locking real trades until after a certain upgrade phase.

This is the practical value brought by configurability.

Many strategies fail for very simple reasons: the assumptions supporting the strategies have expired.

Faster execution will only make the system more efficiently repeat old judgments.

EIP-8363 may ultimately not enter Ethereum in its current form. The short-term market impact may also be limited.

It has already indicated one thing.

The crypto market will change prices, but it will also modify issuance policies and protocol rules. These changes will reshape the expected returns of upper-layer products.

A trading system that treats protocol parameters as permanent conditions will eventually discover that its backtesting describes a market that has already disappeared.

When the market begins to rewrite the rules, trading systems need to first rewrite their understanding before placing the next order.